Op amp matching

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
terkio
Posts: 304
Joined: Thu Feb 07, 2019 3:17 am
Contact:

Op amp matching

Post by terkio »

In the design of an integrator circuit whether it is a Deboo or another implementation, a most important op amp characteristic is low Vos.
For a balanced integrator that uses two op amps, the two Vos, if exacly equal, would provide an integrator with 0 Voltage input offset.
Does it make sense to match op amps Vos ?
Can one count on a good match between the two op amps in a dual chip ?
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Op amp matching

Post by mediatechnology »

I can't speak to Vos but my experience looking at Ios, with dual bipolar bipolar non-bias-compensated op amps, is that they match pretty good.

A 5532 with an I bias of 500 nA might be within 10 nA across op amps when used as an INA.
User avatar
terkio
Posts: 304
Joined: Thu Feb 07, 2019 3:17 am
Contact:

Re: Op amp matching

Post by terkio »

An interesting paper about Vos Ios Ibias, how they come in the output offset of a FDA circuit using THS4551.
https://e2e.ti.com/blogs_/b/analogwire/ ... nt-and-why
This shows, on the given example:

Vos is the major contributor onto the output offset.
Next are Ibias and Ios which highly depend of resistors.
_Ios contibutes to output offset more then Ibias. Ios contribution depends of resistor values ( resistors 5 times larger give 5 times more Voos )
_Ibias contribution depends of resistor values and resistor tolerances.

Temperature drift: Same qualitative results.

It would be interesting to know similar results for a balanced circuit using two op amp instead of one as in the example.
It gives insights about the second stage of preamplifiers using two similar circuits to make a balanced output.

With similar Vos Ios and Ibias characteristics. The output offset voltage Voos+ - Voos- would follow the same qualitative results.
Vos major contributor
Ios contribution depending of resistor values
Ibias contribution depending of resistor values and resistor tolerances.
With other dissimilar Vos Ios and Ibias charateristics, the order in which they are important can be different and depend of resistor values.
I guess estimate that within the same op amp technology, results are the same.
I think the results are better thanks to some cancellation from op amp matching.
Who knows how Vos Ios Ibias match about op amps from the same manufacturing batch ?
Who knows when the op amps are in the same chip, as in a dual or quad ?
I think, going balanced gives better results, going balanced with a dual gives best, I guess estimate we gain 10 times better.
EDIT: I later found, my 10 times better statement is all wrong about the Voltage input offset. The LT1002 datasheet shows there is little gain from matching.

Needless to say, I will implement servo integrator circuits using duals like OPA2277 or ADA4522-2.
This paper gives me insights too, about how Vos Ibias Ios come in differential integrator circuits. Differential Deboo or else.

Thermal drift: When the two amps are in a dual chip, the two op amps are at near the same temperature, there should be about no residual drift.
A single op amp is subject to the ambient temperature. With two op amps in a balanced circuit the drift is subject to the temperature difference of the two chips, giving an order of magnitude lower drift. When in the same chip the temperature difference is very small, giving one more order of magnitude lower drift.
Last edited by terkio on Sun Mar 17, 2019 6:30 pm, edited 4 times in total.
User avatar
terkio
Posts: 304
Joined: Thu Feb 07, 2019 3:17 am
Contact:

Re: Op amp matching

Post by terkio »

I found, offset voltage matching does exist:
https://www.analog.com/media/en/technic ... LT1002.pdf
Guaranteed LT1002A 60µV max offset voltage
Guaranteed LT1002A 40µV max offset voltage matching
This data is disapointing, I was expecting much better from matching. :(
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Op amp matching

Post by mediatechnology »

The LT1002 is bias-current-compensated so it may lose some intrinsic matching it might otherwise have.
But its Ibias would be higher.
User avatar
terkio
Posts: 304
Joined: Thu Feb 07, 2019 3:17 am
Contact:

Re: Op amp matching

Post by terkio »

In my quest for low Input Voltage Offset, a key point for an integrator circuit, I lost hope in matching op amps.
What is better than OPA2277 ?
Are they better ones, at low input voltage offset ?
Is chopper op amps the way to go, in audio applications ?
EDIT chopper op amps like: Zero-drift amplifiers, ADA 4522-2, OPA 2189
An article from TI: Auto-zero amplifiers ease the design of high-precision circuits http://www.ti.com/lit/an/slyt204/slyt204.pdf
KMN
Posts: 53
Joined: Sat Jan 20, 2018 2:49 pm

Re: Op amp matching

Post by KMN »

Lower offset but only slightly so...LT1001AM. @ 15uV max, guaranteed, is the lowest Vos off the shelf opamp I know about. If anybody knows of a lower Vos I would be thankful to you if you would share the info. https://www.analog.com/media/en/technic ... LT1001.pdf
KMN
Posts: 53
Joined: Sat Jan 20, 2018 2:49 pm

Re: Op amp matching

Post by KMN »

I just discovered the LMP2021. It has very low offset and it is pretty quiet.

I just couldn't get the choppers quiet in my circuit, though they are also working for me.

http://www.ti.com/lit/ds/symlink/lmp2021.pdf
User avatar
terkio
Posts: 304
Joined: Thu Feb 07, 2019 3:17 am
Contact:

Re: Op amp matching

Post by terkio »

What can be expected with offset compensation with a trimmer ?

For instance considering OPA2277:
Vos 20µV typ, 65µV max
Vos drift 0.15µV / °C
Can I expect with proper trimming a final Vos based on drift only.
Considering à 10°C to 60°C temperature span, if trimmed at 35°C, this would give about 4µV ( 0.15µV x 25 ).

Is a final Vos 4µV feasible or just a dream ?
User avatar
terkio
Posts: 304
Joined: Thu Feb 07, 2019 3:17 am
Contact:

Re: Op amp matching

Post by terkio »

KMN wrote: Wed Mar 20, 2019 12:52 pm I just discovered the LMP2021. It has very low offset and it is pretty quiet.
http://www.ti.com/lit/ds/symlink/lmp2021.pdf
Fantastic Vos. Trouble is, for me, I need +15V -15V op amps.
I just couldn't get the choppers quiet in my circuit, though they are also working for me.
How bad is it ? How to cope with it ?
Post Reply